Former executive vice president and chief technology officer at Firstsource Corp. Dave Carlson is a 23-year IT veteran who's been decorated with more than a dozen awards and distinctions. However, he admits to having a few weaknesses: helping people advance their careers sometimes prematurely, thinking about big-picture IT issues that are sometimes out of the scope of his job and living in Southern California.

Fortunately, after serving as CIO, CTO and CEO at various companies including Ingram Micro and Kmart, he's learned how to channel his desires. Instead of doling out unearned promotions, he mentors employees. He joins committees that let him work on developing industry-spanning standards. Recently he became senior vice president of technology at New York City-based Gemstar-TV Guide International's offices in sunny Pasadena, Calif.

Here he talks about a decision that he does not regard as wrong, but as one with good and bad results: accepting a CIO role at a dotcom startup, Firstsource, which was based in El Segundo, Calif. We talked with Carlson immediately after his stint at Firstsource.

Startup Challenges

CIO: You have worked at large companies such as Kmart and Ingam Micro that have thousands of employees. Late last year you took a CIO post at Firstsource, a startup online infrastructure procurement vendor with 160 employees. After three and a half months, one of Firstsource's creditors pulled its funding, and the company folded. What did your stint at Firstsource teach you about IT culture in a dotcom?

Carlson: That it's not that different.

Entering Firstsource was quite analogous to my going into Kmart, which was 20 times larger. At Kmart there were two factions in the IT department. I took over one of them my first year and forged strong relationships with the other faction. After a year I took over all of IT and blended them together. It worked because for that year I made sure we viewed the sharks as Sears, Wal-Mart and Target, and not the group on the other side that happens to be managing the mainframe.

Similarly, within Firstsource there was a group of people working on new things that was a third of the IT staff, and the rest of the staff that made up another group was managing the stuff that was already in place. They were not at all involved in the development of the new stuff, though they knew most about what the current activities were. The new things that were being developed desperately needed their input. So, like at Kmart, I had to get cooperative environments established through mutual respect.

What were your core responsibilities at Firstsource?

I was responsible for all IT operations and application development. Nine people reported directly to me. When I started out there were two direct reports, but I was uncomfortable with that. So within four weeks I had nine. I wanted more because I tend to depend on people a lot and think multiple heads are better than one.

How was managing Firstsource's 60-man IT department different from managing the crew of 1,200 you had at Kmart?

I don't know that the challenge was a lot different for me. When you get above the point where you can personally supervise everybody that works for you - that's maybe 10 - from that point on it has to do with one's ability to manage, motivate, lead and make sure you become the advocate for the people who really get the work done. You just have to make absolutely sure you surround yourself with people who are responsible, technically competent and well motivated.

Were Firstsource's processes markedly different than those at a larger company?

No. What I saw at Firstsource - the lack of processes that support the company's objectives - was similar to what I have seen in many places. Most places are pretty process weak. At Firstsource I instituted processes and budget controls in the same way I did at Kmart and Ingram.

What kinds of projects did you work on at Firstsource?

We were working on replacing the operation environment, which was all Microsoft ASP, with a new Java-based environment that would give us a little more scaleability and some flexibility. We were in the midst of reviewing the differences between the two environments when we closed shop.

Do you have any advice for someone who wants to be the CIO of a small company or dotcom startup?

Certainly, you have to be cautious. But if you're too cautious, you end up with a 20-year career that's the same year 20 times. And if you have that kind of personality, you shouldn't be in technology.

There was some risk for me going to Firstsource. But I don't regret it. I was able to build some exciting things with both old and new technology, do it with people I liked, and do it with people who had high integrity. It didn't hurt me because now I have an extraordinary opportunity [at Gemstar-TV Guide International].

The general principles of managing IT don't change - you have to stay on top of the technology, know what you don't know and get really good people to help. You still have to relate to them, have enough respect so that they know that even if there's disagreement they are still a contributor.

What keeps you up at night?

The complexity of the infrastructure.

There are elements in the infrastructure on which the operation is totally, absolutely dependent. There may be only two people in the world who know that little piece of code that runs that.

Two years ago, Ingram Micro was down for eight hours worldwide. The reason it took us eight hours to find it was the complex global network that spans from mainland China to Tucson, Ariz. Turns out that the person who set the defaults in a communications processor set 16 out of 19 and missed three. That happened to be the three that handled all of the United States and most of Europe.

How I Got Here

You spent more than a decade pursuing three degrees in math and engineering. While in school, did you think that you wanted to end up in a technology administrative role?

As an undergraduate in the early '60s I knew that I really liked being around computers and what they could do. I'm kind of obsessive-compulsive, so I managed to get the syntax right in early programs and did very well. That influenced me forever. But I didn't know where I'd apply it.

After you received your PhD in industrial and operations engineering, you landed a job at Allied Supermarkets that was essentially a CIO role. Did your theoretical studies prepare you for business?

No, not necessarily. That's why I went back and took more finance courses.

My boss at Allied, Pat Murphy, challenged me to help him manage inventory in a remote grocery store company. To prepare, he encouraged me to take some additional courses in financial policy.

Before that time, I probably didn't know a balance sheet from an income statement, and certainly had no sense of how capital was leveraged, how to work through a source and use-of-funds statement, so that has been really invaluable.

Throughout your career, what was the most onerous project that you learned the most from?

When I was brought into Kmart [in 1985] they had five different point-of-sale environments. Within two years, we came up with an architecture that allowed all scanning stores in the United States and Canada to be implemented on exactly the same hardware platform. We saved about US40,000 per store with that architecture and saved US80 million to US100 million. That was 15 years ago, when US80 million to US100 million meant something.

If you could change something you did early in your career, what would it be?

I left Allied Supermarkets in 1978 because I got worried about the financial health of the company. I was learning a tremendous amount and had great colleagues. I left in February after three years, and they went bankrupt in November.

I wouldn't want to go back there just to live through a bankruptcy. But I think by staying there a little longer, I probably could have learned a little bit more from Pat Murphy.

What was the biggest mistake you made?

There have been three occasions over my career where I allowed someone to be in a management position, either directly working for me or one step down, longer than they should have. They had taken the position with my encouragement and the implicit agreement that I was going to help them be successful. In every case I spent a lot of time trying to get the corrective action and it just didn't come.

What were you forced to learn on the job that you did not expect to have to learn?

I really didn't think it would be so difficult to get non-technical management, both peers and superiors, to grasp the potential of what the technology could bring to the organisation.

Was there anything you thought was incredibly important starting out that turned out to be not important at all?

I thought it was important to be able to code Cobol like a whiz.

But the other skills are what matter, the human skills. Sometimes it is bringing somebody up very short by saying, "This behaviour is unacceptable." In other cases, it is saying, "What is going wrong? I am concerned about you. Your performance has deteriorated in the last three months. I know something is going on. You don't have to tell me. I am not insisting at all. But get to somebody who can help you." I have certainly had more than my share of those conversations. I didn't know how important they would end up being.

Learning About Management

What have you learned about managing people?

When you start out as a manager, you think your subject-matter knowledge is the most important ingredient in your effectiveness as a manager. It's not. As the CIO of Kmart, the subject-matter expertise probably was no more than 10 percent or 15 percent relevant on an average day. When we talked about making a particular decision to buy - what the architecture was for the master system within the store - then my level of subject-matter expertise on that exact issue went up to 80 percent.

What career moves helped you become a better CIO?

Before I became the CIO of Kmart, I spent a year and a half running a small company that produced point-of-sale equipment. By the time I got to Kmart, I could look at a point-of-sale terminal from any vendor, and I could know what their manufacturing costs were. So we were able to leverage those companies a lot during the time I was there, partially because I really understood the manufacturing process.

As I look back at what kind of ingredients there were to successful career moves, a lot of them had to do with all of the men at this point who would be considered vernacular mentors-both on the technical side, by the way, and on the business side.

Early in a CIO's career, do you think mentors are important?

Yes. They were for me.

There was a mentor of mine named Harvey Wiseber who was executive vice president of a small supermarket chain. Harvey loved to teach and to get to the bottom of things, such as why sales of produce were off. I spent a huge amount of time with what was called "walking the store" with Harvey. He was a tremendously engaged guy, and he wanted me to understand the supermarket business.

There may be a point, though, where you pass out of a phase where mentors become the most influential. That's when highly respected colleagues become the source of not just inspiration, but knowledge and somebody to talk to about what's going on.

What if you don't have such coworkers? Should the up-and-coming CIO's join committees?

I think so. I have found an extraordinary enjoyment in a couple of committees.

A friend of mine told me once, "The trick to management is hiring people with the right demons." That has been a very good phrase in my career. But another way is to surround yourself with professional colleagues who have the right demons. And you can find them in committees.

A guy I know, Tom English, has the, "I want to keep up on everything that's going on that may be important to me" demon. I try to sit next to him at committee meetings, just so I can pick his brain on during the coffee break.

Is it a challenge to manage smart people?

I think there's a point in becoming a manager - and this is often a difficult point for first-time supervisors - when they realise that they can no longer do the jobs that they manage as well as the people who work for them can do. At one time, I was a really, really good assembler code programmer, a Fortran programmer. When I got into management, I had to give that up. You therefore have to make sure that the people you have around you, first as first-line employees and then second as supervision or management, are able to manage down to the high-quality level.

If you look at the hundreds of professionals who worked under me at Kmart, I might have been able to do only a few of their jobs. I may have been able to install a Windows 95 upgrade, but don't let me anywhere near the client server LAN environment, a router or the mainframe.

What's the single biggest lesson you'd pass on to a new CIO?

Recognise that the technology component of the job is always important, but gets less so over time. The human component grows much faster and is much more challenging and potentially rewarding.

The Technology End

Is there any technology that you invested money in that you consider a waste?

I brought in an extraordinarily innovative piece of equipment to Kmart that went in probably three years earlier than it should have. The merchants did not take full advantage of the several million dollars that were spent on it.

My vision, that the technology would revolutionise the way we merchandise the stores, was right. Wal-Mart has demonstrated because it adopted this same technology, that it can do an astoundingly good job of merchandising store-by-store items.

Kmart wasn't ready. It could have generated, with the cooperation of the merchants, a far greater return.

I should have demanded that technology should have at least 100 percent payback in a year, and we would not be content with a 16 percent or 17 percent internal rate of return. If I did I would have had to wait a couple of years.

How important is it for a CIO to be able get in the trenches, roll up his sleeves and get his hands dirty?

When necessary, it's immensely important. That's why I think that knowledge I gained studying mathematics still allows me to listen to arguments and debates about what's going on, what should go on, and detect what seems to be relevant, what is extraneous, what's really important, what's not so important.

That's important because technical people often engage in religious wars. At Kmart we made some important decisions, and they were greatly influenced by people who would sit in a room and debate these things with myself and a couple of highly regarded managers who worked for me, listening in and then saying, "OK, I appreciate all the opinions. If there are any holes in the information, we'll go back and take a look at it, but now a few of us are going to get together and distil it all into the essential ingredients for a decision."